SCHOOL OF MECHANICAL ENGINEERING SEMINAR
Wednesday, October 27, 2021 at 14:00
Wolfson Building of Mechanical Engineering, Room 206
OPTIMAL KINEMATIC DESIGN OF A MANIPULATOR FOR EARLY DETECTION OF STRESSES IN GREENHOUSE CROPS
Tamir Mhabary
Mcs Amiram Moshaiov
This thesis is part of a project dealing with the development of an agricultural robot for early disease detection in greenhouse crops. The robot, mounted with a robotic manipulator arm that will carry the sensors designed for early disease detection, will autonomously navigate through the greenhouse. Detection of diseases, during the growing and harvesting stage, which is conducted by periodic inspection, is essential to prevent the spread of the disease and to significantly reduce the damage to the crop. Today, the monitoring is carried out manually. Due to manpower constraints, it involves a high inspection cost, time and provides insufficient sampling frequency, which may result in pest and disease outbreaks and unnecessary use of pesticides and sprays.
This study aims to find the optimal kinematic design of the manipulator for this task, which will carry the sensors. To find the optimal kinematic design, the research uses methods from the world of multi-objective optimization and evolutionary algorithms.
To solve this optimization problem, existing tools were used, and new tools were developed - both in the field of robotics and in the field of multi-objective optimization.
The main contribution of the thesis is the development of an evolutionary set-based concept search and optimization method that will enable the selection and adaption of robotic arms for performing the agricultural task. The suggested method is also expected to be useful for other tasks which require an automation process in agriculture and other industries. In addition, the approach taken here might be suitable for solving other multi-objective optimization problems.
